Virtual Environment and Keeping Up With Change

                                                                                         

The rapid growth of data center virtualization and the adoption of service-oriented architectures have resulted in a dynamic and fluid enterprise IT infrastructure.

Keeping up with changes with your systems management software can be the challenge. There are tools that are good at understanding infrastructure availability, but they lack visibility into the behavior of virtual resources and application performance.. To truly scale your virtual and physical data center environments, automation is almost necessary.

Optimizing the virtual infrastructure and eventually automating resource allocation will require systems administrators to have a better understanding of virtual machine (VM) behavior and the ability to correlate that understanding with infrastructure performance.

Because VM behavior can be so variable, it will require management tools to be self-learning to adapt constantly to changing behavior patterns.

Deploying a virtual machine (VM) can take minutes or hours as opposed to the days and weeks it takes to provision a physical
machine. VMs are also portable. Intuitive technology lets administrators move live virtual servers within the physical production environment with simple dragging and dropping. Now enterprises are exploring ways to automate this easy movement of VMs to optimize resource allocation and application performance.

IT managers have dozens of management systems that together are supposed to monitor and manage their complex environments.

Historically, IT managers have relied on infrastructure monitoring to assess the availability of infrastructure as well as agents and custom scripts to monitor application performance actively. Systems management tools assist them in provisioning data center resources.

As the environment has scaled, the number of alerts and false positives—despite efforts to correlate alerts across technology
domains—has become unmanageable.

The rapid adoption of service-oriented architectures (SOAs) and the explosive growth of virtualization are rapidly breaking down the
connection between infrastructure and application performance. It has always been difficult, but it is now nearly impossible to infer application performance from infrastructure monitoring. With so many moving parts, IT systems can overwhelm existing management tools with thousands of meaningless performance and availability alerts. Sifting through the noise has always been difficult, but in the age of virtualization, it is humanly impossible and creates many unique challenges.

Virtualization will be the standard in the data center, and critical applications will rely on it. Unpredictable performance, underutilized capacity and poorly integrated management will not only be a problem for IT, it will also pose significant risks for the business.

Infrastructure costs will still be too high and will hurt profitability. User productivity will suffer from poor application performance. Users won’t even adopt new application rollouts if the performance is unpredictable. Performance issues can last longer and affect more users. All of this points to the need for a new approach to managing and optimizing performance in virtual environments.

The key descriptive word is unpredictable. IT is supporting increasingly complex and unpredictable user and technology
interactions. Without the restrictions imposed by siloed, proprietary
infrastructure platforms, performance has become difficult to predict.

To restore predictability and bring performance consistency to virtual environments, management needs to adapt in the following ways:

• Self-learning capability: Heuristics and behavior analysis have been used in the IT security realm for years. Real-time behavior analysis provides the same benefit of self-learning anomaly detection in the data center. Rather than trying to model constantly changing performance variables, performance management should analyze behavior in real time and correlate infrastructure performance quickly to application performance and vice versa.
• Automated threshold management: As part of the self- learning capability, thresholds should be adaptive. Performance management tools for virtual environment should be able to learn and build behavior profiles for servers, VMs and applications and also adapt thresholds for changing behavior.
• Visibility into individual VM and system behavior: There are so many moving parts of a virtual infrastructure that it can be nearly impossible to isolate the cause of an application performance issue.
It’s critical to have visibility into the health of each VM and the health of the overall system. It’s also important to have multiple contexts (e.g., knowing the relative performance of a VM to its host VMware ESX Server as well as the entire pool of resources).
• Proactive capacity planning: Performance management tools need to offer resource allocation and capacity planning before deployment as well as in production. The value of virtualization is flexibility and resource optimization. Tools that can deliver that optimization from the onset are the most valuable.

Enterprise systems administrators will increasingly rely on behavior-based tools that can learn and adapt to changing user demands and IT resources. Automation and dynamic resource allocation relies on this fundamental understanding of IT supply and business demand.

Systems management vendors will need to incorporate this capability quickly. It will facilitate adoption of virtualization if they can help to better manage VM sprawl and make performance management more intuitive and efficient. In the next 3 to 5 years, enterprise data centers will continue their inexorable march toward utility computing and act more and more as an internal service provider. As systems become more open, IT will eventually need to compete with external providers of computing power and application delivery. Self-learning performance management could be the competitive advantage that enables IT to keep control over IT resources, tune services more precisely and meet stringent service-level requirements.

About the author:
Stephen J. Richards has 25 years experience in Data Management and Information Technology. This information is provided as a public service by Neon Enterprise Software, a leading provider of IMS outsourcing. For more information, please visit http://www.neonesoft.com.


Article Source: http://www.Free-Articles-Zone.com